Blogs

100KM

openFoam

Posted by crystal garris on April 28, 2014 at 6:46pm

I wanted to share my experience with openFoam. 

techwing_01.png?w=871

I am using openFoam as part of the free linux distribution CAElinux. CAElinux makes it a snap to get a copy of openFoam up and running. CAElinux also comes with a plethora of open source CAE packages. After going through many of the available tutorials, I can now import my own geometries and run a case. I still have much to learn as openFoam is extremely powerful and flexible. The interface is mostly through setup files so it is not very intuitive or point and click. 

mesh.jpg?w=800

the first step is to produce a computational mesh. this is a screenshot of gmsh an open source meshing program for mesh generation. it is very tricky adjusting the mesh densities. set things a little too fine and you can end up with millions of entities and take days to complete. set it too low and your computations will not be as accurate. I can successfully generate all the meshes required but still working on getting from gmsh to openFoam. 

After the mesh is generated( I used snappyhexmesh for the photo at the top) you must configure the dictionaries and folders in your case file ( I modded the motorBike tutorial case folder). After this is all set up and the code in the dictionary files makes sense, a quick "./Allrun' in the terminal to execute the run script. After a while the case will be done running and you can enter "paraFoam' to open up paraview with your newly created simulation data. paraview has a nice GUI interface and does not take long to produce some nice pictures of the 3D pressure and velocity fields of the airflow around the model. 

I still have much to learn before I can start making highly accurate performance predictions with openFoam, but I know I will get there. it is great to have access to such a powerful and open CFD package

Tags: cfd, aerodynamics, openfoam
E-mail me when people leave their comments –
Follow

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • João August 12, 2014 at 1:00pm

    Dear Wayne,

    I've been trying to simulate a similar scenario of yours. In this meanwhile I already found interesting material, like this AeroFoam solver and its examples: https://www.aero.polimi.it/freecase/?OpenFOAM_%2B_Code_Aster:Aerody...

    But, setting up the solver itself is already very challenging. Do you think you could provide the files for your simulation scenario? My objective is to extract pressure and velocity data from the wing surface in order to detect the separation boundary layer using a sensor network (which is my research topic itself).

    Since fluid mechanics is not my main field, any help would be very important.

    Thanks for your attention.

    João

  • zheng jishi May 3, 2014 at 6:56pm

    hi,wayne,

        How can you make the geometry file?The condition for the OF is the surface of the geometry is hermitic.How can you do the work?

  • 100KM
    crystal garris May 2, 2014 at 12:17pm

    I start my designs with XFLR5. It is great for airfoil selection and plane-form design. For further aerodynamic refinement, you need something much more powerful.

  • Martin April 30, 2014 at 11:28pm

     VLM is a kind of 2.5 D analysis. It  takes 2D airfoil profiles and interpolates them into a plane form. OpenFoam is a 3D Field Operation And Minipulation program. In other words openFoam lets you calculate FEs interactions in any dimension. VLM can not calculate how any 2 surfaces interact. For instance if you want to take a close look at the airflow around the root of a wing where it joins the fuselage, VLM will not do that.

    Take a look at this thread --> http://www.rcgroups.com/forums/showthread.php?t=1751910

    You might be underestimating VLM. It might not do wing-body intersections well, but I really doubt that CFD will do better at low Reynolds.

  • MR60
    Hugues April 30, 2014 at 3:35pm

    Thx. This looks hugely complex! I'm amazed people can even program such a software suit. Too bad there is so much to learn and so little time.

  • 100KM
    crystal garris April 30, 2014 at 8:08am

    @ Martin

     VLM is a kind of 2.5 D analysis. It  takes 2D airfoil profiles and interpolates them into a plane form. OpenFoam is a 3D Field Operation And Minipulation program. In other words openFoam lets you calculate FEs interactions in any dimension. VLM can not calculate how any 2 surfaces interact. For instance if you want to take a close look at the airflow around the root of a wing where it joins the fuselage, VLM will not do that.

    @Hugues,

     OpenFoam is a CFD/FE suit that allows you to set up various kinds of simulations. everything from aerodynamics to stress analysis to DNS simulation and much more. It allows you to run many different kinds of preexisting solvers or you can even program your own. I use it to create a virtual wind tunnel which I can place any arbitrary shape inside and start calculating things like lift and drag.

  • MR60
    Hugues April 30, 2014 at 4:17am

    What is openFoam for dummies (like me) ?

  • Martin April 29, 2014 at 9:46pm

    What is your motivation for using CFD for performance analysis (instead of VLM)?

  • 100KM
    crystal garris April 29, 2014 at 6:56am
    @ andrew
    I have some tweaking to do to get good accuracy. The scale is way off and the mesh needs to be tighter in some areas.
    @ rctech
    Sure! It is cool looking. I'll have more to show.

    @F1P
    The picture at the top is paraview.
  • F1P April 29, 2014 at 5:53am

    U dont use paraview?

This reply was deleted.